Gambling for redemption: Consequential inconsistency in dynamic decision making
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Planning under uncertainty is vital for effective decision-making, but human planners systematically violate normative models by allowing irrelevant past events to distort future choices—a phenomenon termed consequential inconsistency (CI). Despite evidence of CI, little is known about its underlying cognitive mechanisms. Across three experiments, we documented robust evidence for CI—individuals were more likely to gamble following an unfortunate event than after a fortunate one—despite facing consequentially identical prospects. Using computational modeling to test three competing accounts, we found that for most individuals CI is not primarily a failure of probability perception—as posited by the gambler’s fallacy—or utility calculation—as posited by the reference-point shift model—but rather a goal-driven desire to "gamble for redemption," as formalized by the aspiration-level model. These findings elucidate the psychological processes involved in planning, while also revealing the cognitive underpinnings of individual differences in dynamic decision-making.